By the way you said in your hypothesis that p has to be odd, implying that this does not work for p=2. It does in fact work for p=2.
Also, the proof of the hypothesis is far more interesting than the proof of your statement, which is practically a corollary of the hypothesis.